In 30 years of walk-forward data, the dominant factor shifted several times and the magnitude of the shifts were frequent.
The 10 principal components of the current training window, ordered by variance explained. Together they capture 87.1% of variance in the 16-indicator dataset.

| Label | Variance | Cumulative | Top features |
|---|---|---|---|
| PC1 - Equity stress | 22.1 | 22.1 | SPX_200d_ma_spread (−), SPX (−), BAA10Y (+), BAA10Y_trend (+) |
| PC2 - Economic activity | 13.1 | 35.2 | MANEMP (+), SPX_from_52w_high (+), PERMIT (+), BAA10Y_trend (+) |
| PC3 - Rate levels | 9.9 | 45.1 | DTB3 (+), DGS10 (+), RVOL_trend (+), SPX_RVOL21 (+) |
| PC4 - Curve shape | 8.9 | 54 | T10Y2Y (+), DCOILWTICO (+), RVOL_trend (+) |
| PC5 - Dollar / commodity | 7.2 | 61.2 | DXY (+), DCOILWTICO (−), ICSA (+), MICH (−) |
| PC6 - Consumer sentiment | 6.3 | 67.5 | MICH (+), ICSA (+), MANEMP (+) |
| PC7 - Claims / dollar | 5.4 | 73 | ICSA (+), DXY (−), MANEMP (+) |
| PC8 - Housing / volatility | 5.1 | 78.1 | PERMIT (+), T10Y2Y (+), RVOL_trend (−) |
| PC9 - Commodity / credit | 4.9 | 82.9 | DCOILWTICO (+), DTB3 (+), BAA10Y_trend (+) |
| PC10 - Short rates | 4.2 | 87.1 | DTB3 (+), MANEMP (−), DXY (−) |
Important disclosures
The PCA factor analysis presented on this page is provided for informational and research purposes only. It does not constitute investment advice, a solicitation, or a recommendation to buy or sell any security or financial instrument.
Factor analysis results are based on historical data and walk-forward methodology. Past factor structure does not indicate future factor structure. The covariance structure of macro indicators changes over time and may shift materially from the patterns shown.
AppliedComplexity Research makes no representations as to the accuracy or completeness of the information presented. For informational use only.